Software algorithms are not neutral. They embody the thinking of the team that created them. Even if that team has explored a lot of edge cases, there’s no way to be NP-complete on exploring all edge cases when you’re interacting with a human. https://bit.ly/3a58X34
-
Take a look at Cathy O'Neil's book Weapons of Math Destruction. The problem is in the algorithms, the data, and the teams who aren't aware of their assumptions.
Haven't read it, will put it on the list! I'm on an AI reading binge. More recs welcome.
